The Whitehall II study of 10,308 male and female civil servants aged 35-55 years at entry (1985-1988), was established to examine the role of specific psychosocial, lifestyle, biochemical and physiological factors as possible explanations of these inequalities. True age related changes in these exposures, or cumulative exposure measured longitudinally, are hypothesized to predict changes in SES differences in health with age. At the 10-year follow up of the cohort, NIA support funded collection of data to repeat outcome measures of health functioning, cognitive functioning, components of the metabolic syndrome and ApoE genotyping. This application requests funding to analyze the data collected to date and to contribute to specific elements of the 15-year follow-up of the cohort. This funding will enable the investigators to accumulate more endpoints and track health functioning into older age, relate them to early life and mid-life exposures, and thereby allow us to establish psychosocial and biological pathways of disease and health inequalities. The aims of the application are: (1) To describe and explain patterns of change with age in health status in relation to SES; (2) To determine if the gradient in health functioning differs from pre-retirement to retirement; (3) To examine the relationship between SES and change in cognitive function with age; (4) To investigate specific biological pathways linking SES by examining the causes and consequences of their change with age. The Whitehall II study is uniquely poised to address these questions, offering: civil service grade as an excellent measure of SES; longitudinal design with participants comparatively young at entry allowing the detection of antecedents of change; repeated measures of exposures; a wide range of exposure data; substantial power to detect age-related change, and its interaction with SES; wide range of health outcomes including health and cognitive functioning, components of the metabolic syndrome, mortality, non-fatal diagnoses and sickness absence.